如何使用keras获知WGAN中生成的图像的标签

import keras
from keras.models import Sequential
from keras.layers import Dense, Reshape, Flatten
from keras.optimizers import RMSprop

def build_generator():
    generator = Sequential()
    # 添加生成器网络层
    ...
    return generator

def build_discriminator():
    discriminator = Sequential()
    # 添加判别器网络层
    ...
    return discriminator

def build_wgan(generator, discriminator):
    wgan = Sequential()
    wgan.add(generator)
    wgan.add(discriminator)
    return wgan

generator = build_generator()
discriminator = build_discriminator()
wgan = build_wgan(generator, discriminator)
wgan.compile(optimizer=RMSprop(lr=0.00005), loss=wasserstein_loss)

for epoch in range(num_epochs):
    # 生成真实图像和噪声数据
    real_images = ...
    noise = ...
    
    # 生成假图像
    fake_images = generator.predict(noise)
    
    # 训练判别器
    discriminator_loss_real = discriminator.train_on_batch(real_images, real_labels)
    discriminator_loss_fake = discriminator.train_on_batch(fake_images, fake_labels)
    
    # 固定判别器的权重，训练生成器
    discriminator.trainable = False
    generator_loss = wgan.train_on_batch(noise, real_labels)
    discriminator.trainable = True

noise = ...
generated_images = generator.predict(noise)
labels = discriminator.predict(generated_images)